GLOBAL EPIDEMIOLOGICAL DATASET ON TUBERCULOSIS INCIDENCE WITH HEALTH AND SOCIO-ECONOMIC CORRELATES (2000–2022)¶
In [19]:
import pandas as pd
In [21]:
df=pd.read_csv(r"TB 5000.csv")
How does TB incidence correlate with HIV co-infection rates across different age groups and genders?¶
In [83]:
import pandas as pd
import matplotlib.pyplot as plt
# Group data
age_gender_group = df.groupby(['Age_Group', 'Gender']).agg({
'TB_Incidence_Rate': 'mean',
'HIV_CoInfection_Rate': 'mean'
}).reset_index()
# Pivot for visualization
pivot_df = age_gender_group.pivot(index='Age_Group', columns='Gender',
values=['TB_Incidence_Rate', 'HIV_CoInfection_Rate'])
# Plot
fig, ax = plt.subplots(figsize=(12, 6))
pivot_df['TB_Incidence_Rate'].plot(kind='bar', ax=ax, position=1, width=0.4, color=['skyblue', 'salmon'])
pivot_df['HIV_CoInfection_Rate'].plot(kind='bar', ax=ax, position=0, width=0.4, color=['navy', 'darkred'])
ax.set_title('TB Incidence vs. HIV Co-Infection Rates by Age and Gender')
ax.set_xlabel('Age Group')
ax.set_ylabel('Rate per 100,000')
ax.legend(['Female (TB)', 'Male (TB)', 'Female (HIV)', 'Male (HIV)'])
plt.xticks(rotation=45)
plt.savefig("TB incidence and mortality rate.png")
plt.show()
What is the relationship between BCG vaccination coverage and multidrug-resistant TB rates in high-burden countries?¶
In [87]:
# Filter data for BCG coverage > 0
bcg_mdr = df[df['BCG_Vaccination_Coverage'] > 0]
# Plot
plt.figure(figsize=(10, 6))
plt.scatter(bcg_mdr['BCG_Vaccination_Coverage'], bcg_mdr['Multidrug_Resistance_Rate'],
c=bcg_mdr['TB_Incidence_Rate'], cmap='viridis', alpha=0.7)
plt.colorbar(label='TB Incidence Rate')
plt.title('BCG Coverage vs. Multidrug-Resistant TB Rates')
plt.xlabel('BCG Vaccination Coverage (%)')
plt.ylabel('Multidrug-Resistant TB Rate (%)')
plt.grid(True)
plt.savefig("BCG vaccination coverage.png")
plt.show()
Do countries with higher primary healthcare access exhibit improved TB treatment success rates?¶
In [89]:
import seaborn as sns
plt.figure(figsize=(10, 6))
sns.regplot(x=df['Primary_Healthcare_Access'], y=df['Treatment_Success_Rate'],
scatter_kws={'alpha':0.5}, line_kws={'color':'red'})
plt.title('Primary Healthcare Access vs. TB Treatment Success')
plt.xlabel('Primary Healthcare Access (%)')
plt.ylabel('Treatment Success Rate (%)')
plt.savefig("TB treatment success rate.png")
plt.show()
How does TB mortality vary with urban versus rural population distribution?¶
In [91]:
df['Urban_Rural'] = df['Urban_Population_Percentage'].apply(lambda x: 'Urban' if x >= 50 else 'Rural')
plt.figure(figsize=(8, 6))
sns.boxplot(x='Urban_Rural', y='Mortality_Rate', hue='Urban_Rural',data=df, palette='Set2',legend=False)
plt.title('TB Mortality Rate: Urban vs. Rural Populations')
plt.xlabel('Population Type')
plt.ylabel('Mortality Rate (%)')
plt.savefig("TB mortality variation.png")
plt.show()
Are smoking and alcohol consumption significant independent predictors of TB incidence after controlling for GDP and healthcare spending?¶
In [93]:
# Correlation matrix
corr_matrix = df[['TB_Incidence_Rate', 'Smoking_Rate', 'Alcohol_Consumption', 'GDP_Per_Capita']].corr()
plt.figure(figsize=(8, 6))
sns.heatmap(corr_matrix, annot=True, cmap='coolwarm', vmin=-1, vmax=1)
plt.title('Correlation Matrix: TB Incidence and Risk Factors')
plt.savefig("TB incidence after control for GDP and healthcare spending.png")
plt.show()
Does GDP per capita moderate the effectiveness of TB screening programs in reducing incidence rates?¶
In [95]:
# Bin GDP into tertiles
df['GDP_Tertile'] = pd.qcut(df['GDP_Per_Capita'], q=3, labels=['Low', 'Medium', 'High'])
plt.figure(figsize=(10, 6))
sns.lmplot(x='TB_Screening_Coverage', y='TB_Incidence_Rate', hue='GDP_Tertile',
data=df, height=6, aspect=1.5, markers='o', palette='viridis')
plt.title('TB Screening Effectiveness by GDP Tertile')
plt.xlabel('TB Screening Coverage (%)')
plt.ylabel('TB Incidence Rate (per 100k)')
plt.savefig("GDP per capita.png")
plt.show()
<Figure size 1000x600 with 0 Axes>
How do unemployment rates interact with TB incidence and healthcare expenditure in low-income versus high-income nations?¶
In [97]:
plt.figure(figsize=(12, 8))
plt.scatter(df['Unemployment_Rate'], df['TB_Incidence_Rate'],
s=df['Healthcare_Expenditure_Per_Capita']/50,
c=df['GDP_Per_Capita'], alpha=0.6, cmap='plasma')
plt.colorbar(label='GDP Per Capita')
plt.title('Unemployment vs. TB Incidence (Healthcare Spending as Bubble Size)')
plt.xlabel('Unemployment Rate (%)')
plt.ylabel('TB Incidence Rate (per 100k)')
plt.grid(True)
plt.savefig("TB unemployment rates.png")
plt.show()
What is the long-term trend in TB incidence (2000–2022), and which countries demonstrate statistically significant deviations from global patterns?¶
In [99]:
annual_trend = df.groupby('Year')['TB_Incidence_Rate'].agg(['median', 'std', 'count'])
annual_trend['ci'] = 1.96 * annual_trend['std'] / (annual_trend['count'])
plt.figure(figsize=(12, 6))
plt.plot(annual_trend.index, annual_trend['median'], marker='o', color='darkred')
plt.fill_between(annual_trend.index,
annual_trend['median'] - annual_trend['ci'],
annual_trend['median'] + annual_trend['ci'],
alpha=0.2, color='darkred')
plt.title('Global TB Incidence Trends (2000–2022)')
plt.xlabel('Year')
plt.ylabel('Median TB Incidence Rate (per 100k)')
plt.xticks(range(2000, 2023, 2))
plt.savefig("TB incidence and global deviations in countries.png")
plt.show()
Does increased healthcare expenditure per capita correlate with reduced TB mortality, and does this relationship vary by region?¶
In [101]:
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
# Bin GDP into income groups (customizable thresholds)
df['Income_Group'] = pd.cut(
df['GDP_Per_Capita'],
bins=[0, 10000, 30000, float('inf')],
labels=['Low Income', 'Middle Income', 'High Income']
)
# Plot
plt.figure(figsize=(12, 7))
sns.lmplot(
x='Healthcare_Expenditure_Per_Capita',
y='Mortality_Rate',
hue='Income_Group',
data=df,
height=6,
aspect=1.5,
markers=['o', 's', 'D'],
palette='Set1',
scatter_kws={'alpha': 0.6}
)
plt.title('Healthcare Spending vs. TB Mortality by Income Group')
plt.xlabel('Healthcare Expenditure Per Capita (USD)')
plt.ylabel('TB Mortality Rate (%)')
plt.grid(True)
plt.savefig("Healthcare Spending vs. TB Mortality by Income Group.png")
plt.show()
<Figure size 1200x700 with 0 Axes>
How do gender disparities in TB incidence relate to differences in healthcare access or socio-economic status?¶
In [103]:
gender_stats = df.groupby('Gender').agg({
'Primary_Healthcare_Access': 'mean',
'GDP_Per_Capita': 'mean'
}).reset_index()
fig, ax = plt.subplots(figsize=(10, 6))
gender_stats.plot(x='Gender', y=['Primary_Healthcare_Access', 'GDP_Per_Capita'],
kind='bar', ax=ax, color=['teal', 'gold'])
plt.title('Healthcare Access and GDP by Gender')
plt.xlabel('Gender')
plt.ylabel('Rate (%) / GDP (USD)')
plt.xticks(rotation=0)
plt.savefig("Healthcare Access and GDP by Gender.png")
plt.show()
Are countries with higher BCG vaccination coverage less likely to experience pediatric TB cases (0–14 age group)?¶
In [129]:
# Filter pediatric TB cases
pediatric_df = df[df['Age_Group'] == '0-14']
# Plot
plt.figure(figsize=(10, 6))
sns.regplot(
x='BCG_Vaccination_Coverage',
y='TB_Incidence_Rate',
data=pediatric_df,
scatter_kws={'alpha': 0.6, 'color': 'green'},
line_kws={'color': 'red'}
)
plt.title('BCG Vaccination vs. Pediatric TB Incidence (0–14 Years)')
plt.xlabel('BCG Vaccination Coverage (%)')
plt.ylabel('TB Incidence Rate (per 100k)')
plt.grid(True)
plt.savefig("BCG Vaccination vs. Pediatric TB Incidence (0–14 Years).png")
plt.show()
What demographic factors (age, urbanization, life expectancy) are most predictive of TB treatment success rates?¶
In [131]:
import statsmodels.formula.api as smf
# Regression model
model = smf.ols(
'Treatment_Success_Rate ~ Age_Group + Urban_Population_Percentage + Life_Expectancy',
data=df
).fit()
# Summary and coefficients
print(model.summary())
# Plot coefficients
coeffs = model.params.drop('Intercept')
plt.figure(figsize=(8, 4))
coeffs.plot(kind='bar', color='purple')
plt.title('Impact of Demographic Factors on TB Treatment Success')
plt.ylabel('Coefficient Magnitude')
plt.xticks(rotation=45)
plt.savefig("Impact of Demographic Factors on TB Treatment Success.png")
plt.show()
OLS Regression Results
==================================================================================
Dep. Variable: Treatment_Success_Rate R-squared: 0.001
Model: OLS Adj. R-squared: 0.000
Method: Least Squares F-statistic: 1.199
Date: Tue, 04 Mar 2025 Prob (F-statistic): 0.304
Time: 15:21:40 Log-Likelihood: -19340.
No. Observations: 5000 AIC: 3.869e+04
Df Residuals: 4993 BIC: 3.874e+04
Df Model: 6
Covariance Type: nonrobust
===============================================================================================
coef std err t P>|t| [0.025 0.975]
-----------------------------------------------------------------------------------------------
Intercept 81.7533 0.939 87.090 0.000 79.913 83.594
Age_Group[T.15-24] 0.1323 0.522 0.254 0.800 -0.890 1.155
Age_Group[T.25-54] -0.2775 0.516 -0.538 0.591 -1.289 0.735
Age_Group[T.55-64] -0.5333 0.514 -1.038 0.299 -1.541 0.474
Age_Group[T.65+] -0.4253 0.510 -0.833 0.405 -1.426 0.575
Urban_Population_Percentage 0.0060 0.007 0.844 0.399 -0.008 0.020
Life_Expectancy -0.0257 0.013 -2.042 0.041 -0.050 -0.001
==============================================================================
Omnibus: 4168.552 Durbin-Watson: 2.002
Prob(Omnibus): 0.000 Jarque-Bera (JB): 300.843
Skew: -0.050 Prob(JB): 4.71e-66
Kurtosis: 1.802 Cond. No. 519.
==============================================================================
Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
How does the prevalence of multidrug-resistant TB relate to historical trends in treatment success rates and antibiotic misuse?¶
In [133]:
# Group by year
mdr_treatment = df.groupby('Year').agg({
'Multidrug_Resistance_Rate': 'mean',
'Treatment_Success_Rate': 'mean'
})
# Plot
fig, ax1 = plt.subplots(figsize=(12, 6))
ax1.plot(mdr_treatment.index, mdr_treatment['Multidrug_Resistance_Rate'], color='red', marker='o', label='MDR-TB Rate')
ax1.set_xlabel('Year')
ax1.set_ylabel('MDR-TB Rate (%)', color='red')
ax1.tick_params(axis='y', labelcolor='red')
ax2 = ax1.twinx()
ax2.plot(mdr_treatment.index, mdr_treatment['Treatment_Success_Rate'], color='blue', marker='s', label='Treatment Success')
ax2.set_ylabel('Treatment Success Rate (%)', color='blue')
ax2.tick_params(axis='y', labelcolor='blue')
plt.title('MDR-TB and Treatment Success Trends (2000–2022)')
plt.savefig("")
plt.savefig("multidrug resistance rate.png")
plt.show()
Do countries with higher TB screening coverage achieve faster reductions in incidence compared to those relying on passive case detection?¶
In [135]:
# Categorize screening coverage
df['Screening_Group'] = pd.cut(
df['TB_Screening_Coverage'],
bins=[0, 70, 100],
labels=['Low Screening', 'High Screening']
)
# Group by year and screening group
screening_trend = df.groupby(['Year', 'Screening_Group'],observed=False
)['TB_Incidence_Rate'].median().unstack()
# Plot
plt.figure(figsize=(12, 6))
screening_trend.plot(marker='o')
plt.title('TB Incidence Trends by Screening Coverage Group')
plt.xlabel('Year')
plt.ylabel('Median TB Incidence Rate (per 100k)')
plt.legend(title='Screening Group')
plt.grid(True)
plt.savefig("TB Incidence Trends by Screening Coverage Group.png")
plt.show()
<Figure size 1200x600 with 0 Axes>
What is the combined impact of HIV prevalence and healthcare expenditure on TB mortality across sub-Saharan Africa versus other regions?¶
In [137]:
# Pie chart for TB mortality contribution
total_mortality_ssa = df[ssa_mask]['Mortality_Rate'].sum()
total_mortality_other = df[~ssa_mask]['Mortality_Rate'].sum()
plt.figure(figsize=(8, 6))
plt.pie([total_mortality_ssa, total_mortality_other],
labels=['Sub-Saharan Africa', 'Other Regions'],
autopct='%1.1f%%', colors=['#e6550d', '#3182bd'],
startangle=90, wedgeprops={'edgecolor': 'white'})
plt.title('Proportion of Global TB Mortality by Region')
plt.savefig("Proportion of Global TB Mortality by Region.png")
plt.show()
